Replication package- Fiscal Commons and the Education–Growth Nexus: Fiscally-Mediated Threshold Activation Across Developing Economies

This paper introduces the Fiscally-Mediated Threshold Activation (FMTA) hypothesis: government expenditure constitutes a fiscal commons whose scale determines whether education spending generates positive growth externalities. Below a critical threshold, complementary public inputs are insufficient to activate human capital returns; above it, education spending delivers large, significant growth returns. A formal OLG model nests Galor and Zeira (1993) and Barro (1990) as limiting cases. Hansen's (1999) panel threshold regression across 115 countries (1970–2023) estimates the global threshold at 31.90% of GDP, with above-threshold education returns of 0.466 (p < 0.01) and below-threshold returns indistinguishable from zero. Five regional samples document a threshold-return complementarity: higher estimated thresholds co-occur with lower above-threshold returns, consistent with the structural mechanism derived from the model. Within Sub-Saharan Africa, the CFA franc monetary architecture constrains Francophone fiscal expansion to the neighbourhood of the activation threshold — keeping Francophone country-years predominantly in the inactivated regime — while Anglophone SSA operates above it, providing a natural quasi-experiment in fiscal regime assignment. The replication package includes all Stata do-files, the full 115-country panel dataset (1970–2023), and output tables sufficient to reproduce all reported results. Submitted to the Journal of Economic Growth.本文提出财政中介阈值激活（Fiscally-Mediated Threshold Activation, FMTA）假说：政府支出构成一类财政公共品，其规模决定了教育支出能否产生正向增长外部性。当支出水平低于临界阈值时，配套公共投入不足，无法激活人力资本回报；反之，当支出水平突破该阈值后，教育支出将带来显著且可观的增长回报。 本文构建的形式化世代交叠（Overlapping Generations, OLG）模型将Galor与Zeira（1993）及Barro（1990）的研究作为极限情形纳入其中。通过Hansen（1999）提出的面板阈值回归方法，本文对115个国家1970-2023年的面板数据进行估计，得到全球阈值为GDP的31.90%，阈值以上的教育回报系数为0.466（p < 0.01），阈值以下的教育回报则与零无显著差异。五项区域样本分析验证了阈值与回报的互补性：更高的估计阈值往往伴随更低的阈值以上教育回报，这与模型推导的结构机制一致。在撒哈拉以南非洲（Sub-Saharan Africa, SSA）区域内，非洲金融共同体法郎（CFA franc）货币体系限制了法语区国家的财政扩张空间，使其财政支出水平维持在激活阈值附近，导致法语区国家年度观测值大多处于未激活状态；而英语区撒哈拉以南非洲国家的财政支出则普遍突破阈值，这为财政制度划分提供了自然准实验场景。 本研究的复现包包含所有Stata do文件、完整的115国1970-2023年面板数据集，以及足以复现所有报告结果的产出表格。本文已投稿至《Journal of Economic Growth》。